A New Approach for Image Retrieval Based on Color and Spatial Features
一种新的基于颜色和空间特征的图像检索方法

A new content-based color image retrieval method is proposed in this paper,in which both the color content and the spatial features have been taken into account. Based on the special disposal on the HSV color space,the image invariable,hue is selected as the color character,and then an improved accumulative histogram of the hue is calculated as the color feature. To attain the image's shape and spatial information,the image's H-,S-,and V- component are first transformed into status matrix. Then the state correlogram of the status matrix is calculated as the image's spa- tial distribution information. In matching the sindlarity of images,the Guassian model is used to normalize the differ- ent sub-characters distance. Experiments indicate that this method is great effective in boage's similarity retrieval,and has a higher retrieval-rate.